Over-parameterized nonlinear learning - Gradient descent follows the shortest path
Presenter
February 22, 2019
Keywords:
- nonconvex opt
- theory of deep learning
